The goal of the proposed project is to develop magnetic resonance imaging (MRI) methods for the detection of lipid infiltration in the heart. The main purpose of the project is to improve the diagnosis of a condition known as Arrhythmogenic Right Ventricular Dysplasia (ARVD) a disease characterized by fibrofatty infiltration in the right ventricle that leads to ventricular malfunction and death. The proposed methods are based on obtaining separate lipid and water images of the heart which can then be observed separately or together to visualize lipid infiltration. Preliminary results have shown that significant improvements in the detection of lipid infiltration in the heart are obtained with a double-inversion fast spin-echo method (DIRFSE) where the acquisition of lipid and water k-space data is interleaved. Compared to conventional methods used in the clinic for the detection of lipid infiltration in the heart the interleaved DIR-FSE method yields images with better contrast-to-noise ratio and less artifacts caused by flow and motion. The purpose of this grant is to further improve the interleaved DIR-FSE method, and an alternative method based on the acquisition of gradient and spin-echoes (DIR-GRASE), and to evaluate their performance against the methods currently used in the clinic. The specific aims of the work are: (1) To further improve the DIR-FSE and DIR-GRASE methods. This will include a series of technical developments and optimization of imaging parameters aiming at optimal spatial resolution, lipid/water separation, and noise performance. (2) To evaluate the methods proposed in Aims 1 in a clinical study at a magnetic field of 1.5 T. (3) To extend and evaluate methods at a magnetic field of 3.0 T. The central hypothesis of our work is that a better methodology to detect lipid in the heart can significantly improve the diagnosis of ARVD and other pathologies that are characterized by lipid infiltration.